TrendMiner Use Case
Situation
In the sweet world of sugar production, the heart of the operation beats within the crystallization vessels. The journey to crystalized sugar starts by cutting sugar beets and soaking them in hot water inside the diffusion towers. Next, vacuum and heating processes remove most of the water. This results in a syrup. Finally, the crystals form when the syrup is cooled again and returned to normal pressure.
Goals
- Determine the reason for poor product quality in the crystallization vessels.
- Monitor for anomalies in the agitation and stirring cycles.
- Find the ideal batch cycle time to produce the best quality sugar.
Challenge
Engineers had to overlap different batch profiles to assess the differences and potential root causes.
Approach
- Perform a value-based search to identify periods when the reactors in the vessels ran for longer than the ideal batch cycle time.
- Create a Gantt view of the agitation anomalies and color-code the longer batches to easily distinguish them from shorter batches.
- Place a live monitor tile on a dashboard to continuously track product quality through production.
- Analyze the different behaviors in each vessel to find an ideal batch.
Results
- Using TrendMiner, engineers were able to find a way to distinguish between efficient batches – those that lasted less than 1 hour, 5 minutes – and less efficient batches that were longer.
- This allowed them to see that most of the quality issues happened during the 14 similar events found during the past month.
- They then were able to perform live monitoring of the stirrer speed when it fell out of range, especially when the speed fell below 60% at the pivotal production point (40 liters).
- Finally, engineers were able to pinpoint the lack of consistency during the process in the different reactors and the exact moment when the difference is made during the process.
Value
Engineers were able to minimize steam consumption, prevent sieve clogging, and realize a potential 12% reduction in batch cycle time.
Blog Posts
Building a Foundation for AI & Analytics with an Industrial Data Fabric30 August 2024 - 23:12- TrendMiner’s 2024.R2 Release; A Complete Platform to Facilitate Event Analytics and Operationalize Industrial Data8 August 2024 - 16:38
- Bridging the Organizational Gap in Pursuit of an Augmented Factory30 April 2024 - 20:47
